JMIR Publications

ABSTRACT

Background:

Artificial intelligence (AI) has immense potential to change medical education. However, the integration of AI into medical education is accompanied by substantial challenges. While these barriers have received increasing attention in previous literature, exploration of interaction between these components is rare and insufficient for facilitating AI application in medical education.

Objective:

This study aimed to investigate interdependent and causal relationships among factors influencing the adoption of AI in medical education and develop a structural model among these factors. This model may help institutions and educators identify key leverage points and adopt effective strategies to accelerate implementation of AI tools in medical education.

Methods:

The study employed Total Interpretive Structural Modeling (TISM) as the research design. Nine factors were identified through comprehensive literature review and subsequent expert panel discussions. Semi-structured questionnaires were completed by interviewing 11 clinical educators from different disciplines and institutions in Taiwan. Hierarchical relationships were analyzed, and a MICMAC analysis was performed to classify factors based on their driving and dependence power.

Results:

A total of six levels were determined after partition. The policy and regulation factor and the data privacy and ethics factor were located at the fundamental root level of the TISM model and were classified as driving factors. Factors including technology integration, data source, explainability, and educators' AI cognition and knowledge, occupied the intermediary levels. The software and hardware infrastructure factor was located at the top level of the hierarchy and the learner factor was isolated from other elements.

Conclusions:

The core challenge of integrating AI into medical education may not stem from the technical development of AI, but rather from the policies, regulations and privacy concerns. To improve adoptions of AI in medical education, updating policies and regulations with appropriate privacy and ethics consideration is the top priority. Enhancement of AI competencies of educators and integration of AI practitioners' capabilities are also essential and should be prioritized prior to upgrading software and hardware infrastructure.